【摘要】 针对行波管放大器进行在轨工作状态评估，预测产品未来的工作状态，从而有效保障卫星系统的安全运行。文章通过对某型号行波管放大器的在轨遥测数据进行处理后，采用Informer模型对产品的功率遥测和行波管温度进行了长序列时间预测。模型训练的历史遥测数据长度为240个时间步长，预测目标则为60个时间步长，预测任务共进行了6轮，并使用真实遥测数据进行对比验证。结果显示，模型的预测结果与真实数据表现出一致的趋势和周期性。此模型为行波管放大器的在轨运行提供了一种有效的评估手段。更多还原

【Abstract】 Conducting on-orbit status evaluation of the traveling wave tube amplifiers(TWTAs) and predicting their future working status can effectively ensure the safe operation of the satellite system. By processing the on-orbit telemetry data of the TWTAs, the informer model is employed to perform long-sequence time prediction on the output power of the product and the temperature of the TWTs. The time-step of the historical telemetry data for model training are set to 240, while the prediction target is set to 60 time-step. The prediction task is executed for six rounds, and the real telemetry data is used for comparison. The comparison results show that the prediction results of the model exhibit consistent trends and periodicity with the real data. The model provides an effective evaluation method for the on-orbit operation of the TWTAs.更多还原